INDEX

Explanations

words related to criticism or blaming

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, specifically commas

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

externalActionCode

-0.77

ENGTH

-0.76

TPPStreamerBot

-0.72

ĸļ

-0.69

arthed

-0.66

dinand

-0.62

rine

-0.62

 legion

-0.61

Wil

-0.61

verse

-0.60

POSITIVE LOGITS

imov

0.67

Aut

0.64

 activ

0.64

 Crusher

0.62

hops

0.61

 incub

0.60

ffe

0.60

anim

0.59

 slides

0.59

 Drift

0.59

Activations Density 0.000%

words related to criticism or blaming

punctuation marks, specifically commas

No Known Activations