INDEX

Explanations

locations or geographic regions

oai_token-act-pair · gpt-3.5-turbo

references to geographic locations and specific groups or entities

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

 Features

-0.63

 Scient

-0.61

Dialog

-0.61

 Siber

-0.60

_(

-0.60

Quantity

-0.60

ocide

-0.58

Lv

-0.57

?,

-0.57

clusively

-0.57

POSITIVE LOGITS

 rejoice

0.90

 have

0.85

 reacted

0.84

 were

0.84

 argue

0.84

are

0.82

 weren

0.81

 insist

0.81

 contend

0.78

 knew

0.78

Activations Density 0.560%

locations or geographic regions

references to geographic locations and specific groups or entities

No Comments

No Known Activations