INDEX

Explanations

terms related to journalism and reporters

oai_token-act-pair · gpt-3.5-turbo

references to reporters and journalists

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

BuyableInstoreAndOnline

-0.73

 Defin

-0.69

uras

-0.66

ranged

-0.66

STRUCT

-0.65

 Causes

-0.63

inho

-0.62

emonic

-0.62

ujah

-0.62

ensions

-0.62

POSITIVE LOGITS

 covering

1.10

 interviewed

0.93

 interviewing

0.92

orters

0.88

 journalists

0.87

 Journalists

0.87

 reporter

0.86

 corps

0.85

 reporters

0.83

 reporting

0.83

Activations Density 0.080%

terms related to journalism and reporters

references to reporters and journalists

No Comments

No Known Activations