INDEX

Explanations

terms related to different professions and organizations, especially from military and sports contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to sports events and their associated experiences

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SOURCE

-0.84

CHQ

-0.73

 Ibid

-0.73

Published

-0.72

 UNCLASSIFIED

-0.69

byn

-0.68

Reference

-0.67

source

-0.66

pdf

-0.66

)",

-0.66

POSITIVE LOGITS

juries

0.80

 matchup

0.75

 offseason

0.71

 Stadium

0.71

 SPORTS

0.71

ESPN

0.69

 Sports

0.69

 bolted

0.68

collar

0.68

sports

0.67

Activations Density 1.131%

terms related to different professions and organizations, especially from military and sports contexts

terms related to sports events and their associated experiences

No Comments

No Known Activations