INDEX

Explanations

names of specific individuals, particularly surnames

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names, particularly "Sherman" and "Stanton."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

rium

-0.90

versions

-0.86

 Gutenberg

-0.80

erg

-0.76

rina

-0.76

err

-0.73

Phi

-0.72

estone

-0.69

 Gujar

-0.69

eros

-0.68

POSITIVE LOGITS

 Oaks

1.11

 Sherman

0.95

agements

0.88

stein

0.80

PLIED

0.77

agement

0.71

 Firefly

0.69

mun

0.69

 Dome

0.68

Osw

0.68

Activations Density 0.024%

names of specific individuals, particularly surnames

mentions of specific names, particularly "Sherman" and "Stanton."

No Comments

No Known Activations