INDEX

Explanations

phrases or sentences indicating a topic or focus of discussion

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate discussions about what something is not related to

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

interstitial

-0.92

ãģ®ç

-0.70

 awoken

-0.70

 nonetheless

-0.69

©¶æ¥µ

-0.67

ãĤ¨

-0.66

âĹ¼

-0.66

pez

-0.65

åĮ

-0.64

-0.63

POSITIVE LOGITS

 necessarily

1.03

nor

0.96

 anymore

0.92

 mere

0.79

 malice

0.79

 anything

0.76

 merits

0.75

 merely

0.74

 overtly

0.73

 outright

0.71

Activations Density 0.346%

phrases or sentences indicating a topic or focus of discussion

phrases that indicate discussions about what something is not related to

No Comments

No Known Activations