INDEX

Explanations

the word "coverage"

oai_token-act-pair · gpt-3.5-turbo

references to media coverage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Sabha

-0.74

rir

-0.73

entle

-0.68

raine

-0.65

geries

-0.64

RS

-0.64

gery

-0.61

batch

-0.60

sa

-0.60

 Tale

-0.60

POSITIVE LOGITS

 Coverage

0.91

 coverage

0.86

ages

0.80

verage

0.76

aries

0.72

orship

0.72

abilities

0.72

luster

0.68

 Journalists

0.67

cov

0.66

Activations Density 0.035%

the word "coverage"

references to media coverage

No Comments

No Known Activations