INDEX

Explanations

proper nouns related to politics and media

oai_token-act-pair · gpt-3.5-turbo

references to specific organizations or people associated with political commentary

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

aton

-0.84

fil

-0.82

hog

-0.77

film

-0.73

ands

-0.72

byter

-0.70

elled

-0.68

abilities

-0.66

ouri

-0.66

atted

-0.66

POSITIVE LOGITS

 Leilan

0.81

 ante

0.76

sburgh

0.73

Roz

0.71

 Hots

0.70

apy

0.70

 Thatcher

0.63

azines

0.63

Phoenix

0.61

mopolitan

0.60

Activations Density 0.059%

proper nouns related to politics and media

references to specific organizations or people associated with political commentary

No Comments

No Known Activations