INDEX

Explanations

references to events involving physical locations, particularly related to leaving or departing

oai_token-act-pair · gpt-3.5-turbo

references to locations and settings where actions occur

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Visual

-0.70

Visual

-0.62

Strike

-0.60

ourke

-0.58

 collaborator

-0.58

sports

-0.57

 collaboration

-0.56

 Hann

-0.55

ãģķ

-0.54

 collaborators

-0.54

POSITIVE LOGITS

 prematurely

1.05

 voluntarily

1.01

 abruptly

0.96

 willingly

0.92

 safely

0.89

 altogether

0.88

washer

0.84

 sooner

0.83

enium

0.82

 electronically

0.82

Activations Density 0.198%

references to events involving physical locations, particularly related to leaving or departing

references to locations and settings where actions occur

No Comments

No Known Activations