INDEX

Explanations

words related to legal or criminal actions

oai_token-act-pair · gpt-3.5-turbo

overall positive assessments or evaluations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SPONSORED

-0.63

merce

-0.63

Benz

-0.59

issors

-0.58

rys

-0.58

rity

-0.57

kefeller

-0.57

 weap

-0.56

Quantity

-0.56

ichever

-0.56

POSITIVE LOGITS

onian

0.54

 Logged

0.51

 ãĤµ

0.48

Id

0.48

 Together

0.47

Cod

0.46

?,

0.46

 Enough

0.46

 easily

0.45

 accelerator

0.45

Activations Density 0.000%

words related to legal or criminal actions

overall positive assessments or evaluations

No Comments

No Known Activations