INDEX

Explanations

indirect effects or connections

oai_token-act-pair · gpt-3.5-turbo

references to indirect effects and impacts

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

erk

-0.94

ciating

-0.83

CENT

-0.81

imen

-0.77

ingers

-0.76

ikers

-0.74

 Banner

-0.70

imens

-0.70

 Rampage

-0.70

kar

-0.69

POSITIVE LOGITS

 indirectly

0.85

 indirect

0.84

 reciproc

0.74

ifference

0.69

 subsid

0.69

 beneficial

0.68

 somew

0.68

 sensing

0.68

 reven

0.67

 coupling

0.67

Activations Density 0.018%

indirect effects or connections

references to indirect effects and impacts

No Comments

No Known Activations