INDEX

Explanations

quotes from people combined with a statement or observation

oai_token-act-pair · gpt-3.5-turbo

phrases that include dialogue or speech attribution

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ļéĨĴ

-1.07

ramid

-0.93

displayText

-0.86

¬¼

-0.85

ħĭ

-0.77

 ballot

-0.77

ICAN

-0.76

»Ĵ

-0.74

£ı

-0.74

raped

-0.71

POSITIVE LOGITS

 explains

1.81

 says

1.61

 recalls

1.52

 observes

1.49

 admits

1.46

 remembers

1.39

 concedes

1.34

 recounts

1.32

 agrees

1.31

 laughs

1.30

Activations Density 0.121%

quotes from people combined with a statement or observation

phrases that include dialogue or speech attribution

No Comments

No Known Activations