INDEX

Explanations

news-related phrases or terms, likely related to newsletter subscriptions and news headlines

oai_token-act-pair · gpt-3.5-turbo

references to newsletters and news delivery services

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

etheless

-0.80

soever

-0.78

terday

-0.69

banks

-0.69

ADRA

-0.67

Reviewer

-0.63

ranch

-0.63

BIP

-0.62

 entity

-0.62

portion

-0.61

POSITIVE LOGITS

 FREE

1.02

 Free

0.80

 Updates

0.69

free

0.68

 newsletter

0.67

 sidx

0.67

 Tickets

0.66

 updates

0.66

entary

0.64

 Cycling

0.64

Activations Density 0.070%

news-related phrases or terms, likely related to newsletter subscriptions and news headlines

references to newsletters and news delivery services

No Comments

No Known Activations