INDEX

Explanations

the word "documentary"

oai_token-act-pair · gpt-3.5-turbo

references to documentaries

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

ername

-0.85

 Unic

-0.70

qus

-0.70

reset

-0.70

cffff

-0.70

 keyboards

-0.69

achus

-0.66

etter

-0.65

 combos

-0.64

ebus

-0.64

POSITIVE LOGITS

 documentary

1.87

 documentaries

1.80

 filmmaker

1.65

 filmmakers

1.60

 film

1.55

 filmed

1.45

 filming

1.40

 filmmaking

1.38

 footage

1.37

 films

1.32

Activations Density 0.281%

the word "documentary"

references to documentaries

No Comments

No Known Activations