INDEX

Explanations

statements or events that were broadcasted on television

oai_token-act-pair · gpt-3.5-turbo

references to televised events and broadcasts

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

istries

-0.76

erie

-0.71

hart

-0.71

notations

-0.71

une

-0.71

osit

-0.68

uv

-0.68

ais

-0.67

err

-0.66

ube

-0.65

POSITIVE LOGITS

 televised

1.02

 spectacle

0.81

 announ

0.81

 satell

0.75

 millenn

0.74

ÃįÃį

0.73

soDeliveryDate

0.72

 broadcasts

0.72

 EDITION

0.71

 VIDEOS

0.70

Activations Density 0.022%

statements or events that were broadcasted on television

references to televised events and broadcasts

No Comments

No Known Activations