INDEX

Explanations

references to podcasts or discussions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "talk" 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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

anmar

-0.80

iscal

-0.74

 stocking

-0.71

cher

-0.69

ettel

-0.67

\/\/

-0.64

undai

-0.64

oshenko

-0.63

olson

-0.63

rowing

-0.63

POSITIVE LOGITS

 Talk

0.91

ership

0.83

hran

0.80

edIn

0.78

ebus

0.73

bone

0.73

erness

0.72

IUM

0.71

tion

0.68

Talk

0.68

Activations Density 0.028%

references to podcasts or discussions

instances of the word "talk" and its variations

No Comments

No Known Activations