INDEX

Explanations

news organization names, specifically CNN, and potentially related information or events involving them

oai_token-act-pair · gpt-3.5-turbo

mentions of CNN as a source or reference

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

phant

-0.74

hot

-0.73

ocene

-0.71

vals

-0.70

vol

-0.64

 therap

-0.64

irts

-0.64

mits

-0.63

plates

-0.63

strument

-0.63

POSITIVE LOGITS

Money

1.11

 affiliate

0.97

NOW

0.88

ienne

0.85

ileaks

0.84

CNN

0.82

ial

0.81

pedia

0.80

ILLE

0.79

 anchor

0.79

Activations Density 0.017%

news organization names, specifically CNN, and potentially related information or events involving them

mentions of CNN as a source or reference

No Comments

No Known Activations