INDEX

Explanations

mentions of individuals in positions of authority or power

oai_token-act-pair · gpt-3.5-turbo

phrases indicating presence or existence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

pointers

-0.69

 overfl

-0.65

 milestones

-0.65

 classmates

-0.63

Submit

-0.63

lot

-0.59

dx

-0.59

 ethic

-0.59

pts

-0.58

 qualifiers

-0.58

POSITIVE LOGITS

 charge

1.45

 Charge

1.11

charge

1.11

 office

1.04

 Downing

0.93

effect

0.93

office

0.93

efficiency

0.91

 attendance

0.90

 Baghdad

0.90

Activations Density 0.217%

mentions of individuals in positions of authority or power

phrases indicating presence or existence

No Comments

No Known Activations