INDEX

Explanations

phrases indicating denial or dismissal of something

oai_token-act-pair · gpt-3.5-turbo

references to dismissive or skeptical attitudes toward claims, accusations, and rumors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

engeance

-0.63

 somet

-0.62

 ILCS

-0.61

 Surv

-0.59

acebook

-0.58

 Balanced

-0.57

 pairs

-0.56

omew

-0.56

bucks

-0.56

 interacts

-0.56

POSITIVE LOGITS

 altogether

1.11

 outright

0.99

 entirely

0.82

 dismissing

0.80

 anymore

0.80

esson

0.78

 baseless

0.76

itism

0.76

 alleging

0.75

 portraying

0.73

Activations Density 0.245%

phrases indicating denial or dismissal of something

references to dismissive or skeptical attitudes toward claims, accusations, and rumors

No Comments

No Known Activations