INDEX

Explanations

phrases related to statements made by authorities or officials

oai_token-act-pair · gpt-3.5-turbo

instances of reported speech

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

ptives

-0.69

âĸ¬âĸ¬

-0.67

obyl

-0.66

âĸĳâĸĳ

-0.66

rafted

-0.64

oil

-0.64

Fuck

-0.61

ement

-0.60

Lot

-0.59

olon

-0.59

POSITIVE LOGITS

enz

0.71

doms

0.66

sth

0.64

itud

0.62

 Fargo

0.61

 aides

0.58

 dism

0.58

 indo

0.58

 anecd

0.57

 è£ıè

0.57

Activations Density 0.172%

phrases related to statements made by authorities or officials

instances of reported speech

No Comments

No Known Activations