INDEX

Explanations

mentions of reporters or journalists

oai_token-act-pair · gpt-3.5-turbo

references to journalists or reporters within the text

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

emonic

-0.76

ensions

-0.74

inence

-0.70

oku

-0.70

phal

-0.69

uras

-0.69

ujah

-0.69

ranged

-0.68

inho

-0.68

perm

-0.67

POSITIVE LOGITS

 covering

0.98

 reporter

0.87

 reporters

0.84

 journalists

0.84

orters

0.84

 interviewed

0.84

 Journalists

0.83

 interviewing

0.81

 interviews

0.77

 journalist

0.77

Activations Density 0.059%

mentions of reporters or journalists

references to journalists or reporters within the text

No Comments

No Known Activations