INDEX

Explanations

mentions of flags and associated actions or situations

oai_token-act-pair · gpt-3.5-turbo

references to flags and their significance, particularly in a national context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

enz

-0.89

conom

-0.84

Room

-0.82

 Reviews

-0.82

Db

-0.82

 therapists

-0.81

alysis

-0.81

essions

-0.80

Ther

-0.80

akov

-0.78

POSITIVE LOGITS

 flag

1.89

 flags

1.75

Flag

1.71

 Flag

1.69

flag

1.65

Flags

1.54

 Flags

1.52

 emblem

1.49

 salute

1.46

 mascot

1.45

Activations Density 0.362%

mentions of flags and associated actions or situations

references to flags and their significance, particularly in a national context

No Comments

No Known Activations