INDEX

Explanations

references to something being heavily emphasized or involved in a sentence

oai_token-act-pair · gpt-3.5-turbo

phrases that involve the term "heavily" indicating a significant degree or intensity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

arium

-0.80

iry

-0.75

atu

-0.71

afety

-0.71

vity

-0.70

omore

-0.68

Freedom

-0.68

 Chaser

-0.67

keepers

-0.67

azi

-0.66

POSITIVE LOGITS

 reliant

0.97

 discounted

0.94

 indebted

0.90

 invested

0.88

 redacted

0.86

 armed

0.86

 weighted

0.85

 recruited

0.84

 subsid

0.84

 pregnant

0.83

Activations Density 0.037%

references to something being heavily emphasized or involved in a sentence

phrases that involve the term "heavily" indicating a significant degree or intensity

No Comments

No Known Activations