INDEX

Explanations

examples of technical terminology or jargon, such as abbreviations and specialized terms

oai_token-act-pair · gpt-3.5-turbo

references to external conditions or entities affecting situations

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

 Conversation

-0.65

 unpre

-0.63

 reinforcements

-0.57

 lifes

-0.57

 transport

-0.56

 sleeper

-0.55

 provisional

-0.54

 uproar

-0.54

 tongues

-0.54

 Subtle

-0.54

POSITIVE LOGITS

pecially

1.05

typically

0.86

lev

0.82

-)

0.79

usually

0.79

hing

0.78

.,

0.78

prising

0.78

viously

0.77

including

0.77

Activations Density 0.059%

examples of technical terminology or jargon, such as abbreviations and specialized terms

references to external conditions or entities affecting situations

No Comments

No Known Activations