INDEX

Explanations

words related to television and TV shows

oai_token-act-pair · gpt-3.5-turbo

references to television

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

forth

-0.87

xual

-0.86

ussen

-0.83

ebook

-0.74

iveness

-0.74

currency

-0.73

cies

-0.72

-0.68

theless

-0.68

naires

-0.68

POSITIVE LOGITS

 Trop

1.16

trop

1.11

 Shows

1.00

ILLE

0.92

Guide

0.86

 presenter

0.84

NZ

0.82

 commercials

0.82

LINE

0.81

plays

0.80

Activations Density 0.032%

words related to television and TV shows

references to television

No Comments

No Known Activations