INDEX

Explanations

television-related phrases and terms

oai_token-act-pair · gpt-3.5-turbo

instances of the abbreviation "TV" in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

xual

-0.82

-0.77

ctuary

-0.75

inst

-0.70

cies

-0.69

vasive

-0.68

htaking

-0.68

cy

-0.68

andestine

-0.67

vironment

-0.67

POSITIVE LOGITS

TVs

0.87

levision

0.86

ILLE

0.86

 presenter

0.81

TV

0.81

izoph

0.80

trop

0.79

Guide

0.78

 antennas

0.77

 viewership

0.76

Activations Density 0.016%

television-related phrases and terms

instances of the abbreviation "TV" in the text

No Comments

No Known Activations