INDEX

Explanations

words related to questioning or opposing something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "challenge" and its variations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

otide

-0.84

abet

-0.83

ahon

-0.80

ophe

-0.71

seed

-0.71

gas

-0.69

oiler

-0.67

alez

-0.64

rigan

-0.63

translation

-0.63

POSITIVE LOGITS

enged

0.93

 incumb

0.79

ioned

0.76

 challenging

0.72

chall

0.71

ingly

0.68

 insur

0.68

ively

0.67

 unanswered

0.67

 challengers

0.66

Activations Density 0.037%

words related to questioning or opposing something

instances of the word "challenge" and its variations

No Comments

No Known Activations