INDEX

Explanations

words related to software and technical terms

oai_token-act-pair · gpt-3.5-turbo

references to specific actions or significant events in narratives

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

 challeng

-0.53

sample

-0.44

Interstitial

-0.43

Tokens

-0.41

'"

-0.41

artifacts

-0.40

rul

-0.40

'."

-0.40

 undermin

-0.39

Downloadha

-0.39

POSITIVE LOGITS

ivil

0.47

 Mechdragon

0.43

onis

0.42

irc

0.41

ompl

0.41

igl

0.41

ilyn

0.40

pher

0.39

iac

0.39

 unfocusedRange

0.38

Activations Density 1.544%

words related to software and technical terms

references to specific actions or significant events in narratives

No Comments

No Known Activations