INDEX

Explanations

references to locations or entities with the word "Wood" in them

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Wood."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 masc

-0.78

cheat

-0.75

 aliases

-0.74

 lambda

-0.70

 crore

-0.68

 immunity

-0.66

 invari

-0.65

@@

-0.63

fri

-0.63

å£«

-0.63

POSITIVE LOGITS

 Wood

3.70

Wood

2.65

 Woods

1.90

 wood

1.84

 Woody

1.54

wood

1.54

WOOD

1.50

 Wooden

1.47

 Woodward

1.39

Oak

1.38

Activations Density 0.010%

references to locations or entities with the word "Wood" in them

occurrences of the word "Wood."

No Comments

No Known Activations