INDEX

Explanations

information related to investigations and reports by various agencies

oai_token-act-pair · gpt-3.5-turbo

words and phrases related to investigations and accountability

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

 Democr

-0.86

izons

-0.76

equal

-0.75

avored

-0.73

emon

-0.71

ichever

-0.70

 imag

-0.69

 democracies

-0.69

tarians

-0.69

 afterward

-0.69

POSITIVE LOGITS

 vandalism

1.11

 outage

1.09

 thefts

1.05

 incidents

1.04

 incident

1.01

 breach

1.00

 fentanyl

0.99

 disturbance

0.97

 CCTV

0.95

 leak

0.95

Activations Density 0.773%

information related to investigations and reports by various agencies

words and phrases related to investigations and accountability

No Comments

No Known Activations