INDEX

Explanations

verbs indicating some form of removal, reversal, or suspension

oai_token-act-pair · gpt-3.5-turbo

terms related to the cessation or reversal of actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rics

-0.70

resents

-0.62

acion

-0.61

idth

-0.61

Fit

-0.60

ocl

-0.59

dom

-0.59

sit

-0.58

ipeg

-0.58

Pr

-0.58

POSITIVE LOGITS

by

0.84

 aback

0.73

 entirely

0.69

BY

0.67

livious

0.66

ĸļ

0.65

oing

0.64

 proport

0.64

 accordingly

0.63

 unanimously

0.62

Activations Density 0.262%

verbs indicating some form of removal, reversal, or suspension

terms related to the cessation or reversal of actions or processes

No Comments

No Known Activations