INDEX

Explanations

explicit mentions of past actions or events

oai_token-act-pair · gpt-3.5-turbo

instances of the words "had" and "have"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

oshi

-0.65

owe

-0.65

 Refer

-0.61

andan

-0.61

bia

-0.59

bery

-0.59

 guiName

-0.57

awa

-0.56

Specifically

-0.56

inducing

-0.56

POSITIVE LOGITS

 been

1.00

 gotten

1.00

 nightmares

0.90

 plenty

0.87

 recourse

0.87

gotten

0.84

 gone

0.82

 tons

0.80

 lots

0.78

 trouble

0.77

Activations Density 0.243%

explicit mentions of past actions or events

instances of the words "had" and "have"

No Comments

No Known Activations