INDEX

Explanations

adjectives related to positive qualities like caring, nurturing, strong, fair, broad, thoughtful, inclusive, legal, sober, divisive, repetitive, provocative, kind, perfect, steady, heavier, unsettling, open, specific, and ignorant

oai_token-act-pair · gpt-3.5-turbo

descriptors indicating quality and emotional states, especially in relation to care, fairness, and social issues

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

é¾į

-0.92

osponsors

-0.85

apper

-0.76

channelAvailability

-0.73

anwhile

-0.72

 Parables

-0.70

é¾įå

-0.67

ensibly

-0.65

ipedia

-0.63

indal

-0.63

POSITIVE LOGITS

0.73

 Crimean

0.66

 Commonwealth

0.65

-,

0.63

,...

0.63

istic

0.62

 Sabb

0.61

-'

0.60

0.59

 bast

0.57

Activations Density 0.172%

adjectives related to positive qualities like caring, nurturing, strong, fair, broad, thoughtful, inclusive, legal, sober, divisive, repetitive, provocative, kind, perfect, steady, heavier, unsettling, open, specific, and ignorant

descriptors indicating quality and emotional states, especially in relation to care, fairness, and social issues

No Comments

No Known Activations