INDEX

Explanations

references to military commanders

oai_token-act-pair · gpt-3.5-turbo

references to military or authoritative titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

athe

-0.82

ammed

-0.82

ee

-0.79

ees

-0.79

agi

-0.74

ized

-0.72

ured

-0.70

hews

-0.70

urg

-0.70

ulus

-0.70

POSITIVE LOGITS

hower

0.71

 Kham

0.71

 Dresden

0.67

\\\\\\\\\\\\\\\\

0.65

EStream

0.64

EGIN

0.64

renheit

0.64

BIL

0.64

 Mons

0.64

ptions

0.63

Activations Density 0.157%

references to military commanders

references to military or authoritative titles

No Comments

No Known Activations