INDEX

Explanations

violent actions or emotions

oai_token-act-pair · gpt-3.5-turbo

prepositions indicating location or position

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

rolet

-0.66

idays

-0.61

 WITHOUT

-0.59

 importantly

-0.58

 PLEASE

-0.57

 incumb

-0.57

 released

-0.57

 someday

-0.56

 safely

-0.56

onym

-0.56

POSITIVE LOGITS

 unison

1.37

 front

1.16

 lieu

1.14

animate

1.13

between

1.09

aud

1.05

ordinate

1.03

 accordance

1.00

 favor

1.00

effect

1.00

Activations Density 0.329%

violent actions or emotions

prepositions indicating location or position

No Comments

No Known Activations