INDEX

Explanations

proper nouns associated with people or places

oai_token-act-pair · gpt-3.5-turbo

words and acronyms related to specific locations or entities

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

bats

-0.71

session

-0.61

 leash

-0.60

 Pigs

-0.58

 lifes

-0.57

 behold

-0.55

Reviewer

-0.55

 succ

-0.54

Beat

-0.53

 cough

-0.53

POSITIVE LOGITS

agus

0.77

omal

0.73

iaz

0.72

ãĥĨ

0.70

thodox

0.70

cliffe

0.70

orio

0.69

IUM

0.68

uala

0.67

agin

0.67

Activations Density 0.071%

proper nouns associated with people or places

words and acronyms related to specific locations or entities

No Comments

No Known Activations