INDEX

Explanations

calls to action for signing up for newsletters

oai_token-act-pair · gpt-3.5-turbo

mentions of newsletters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

jury

-0.75

bra

-0.72

phal

-0.71

avery

-0.71

come

-0.69

roo

-0.68

phen

-0.67

had

-0.66

hua

-0.65

governmental

-0.65

POSITIVE LOGITS

 inbox

1.04

 subscriptions

1.00

 subscribers

0.97

 newsletter

0.96

 newsletters

0.93

 Subscribe

0.92

 subscription

0.91

 Seym

0.90

 notifications

0.84

 subscriber

0.79

Activations Density 0.019%

calls to action for signing up for newsletters

mentions of newsletters

No Comments

No Known Activations