INDEX

Explanations

legal and administrative terms or entities

oai_token-act-pair · gpt-3.5-turbo

references to legal entities and organizations

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

Ô

-0.73

DonaldTrump

-0.71

;;;;

-0.69

azeera

-0.66

 Lanka

-0.66

thood

-0.65

 Kard

-0.65

amia

-0.65

odox

-0.64

piring

-0.63

POSITIVE LOGITS

 opted

0.87

 chose

0.84

's

0.77

 consulted

0.76

 asked

0.76

 resorted

0.76

 itself

0.76

 responded

0.76

 stumbled

0.75

 hesitated

0.75

Activations Density 0.480%

legal and administrative terms or entities

references to legal entities and organizations

No Comments

No Known Activations