INDEX

Explanations

terms related to a specific acronym "BB"

oai_token-act-pair · gpt-3.5-turbo

references to "BB" in various contexts

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

izational

-0.85

bourg

-0.84

ional

-0.81

ources

-0.81

perse

-0.79

izations

-0.76

endez

-0.76

angered

-0.74

ainment

-0.73

pora

-0.72

POSITIVE LOGITS

BB

0.98

zeb

0.79

ZZ

0.77

ATH

0.76

FW

0.75

ooth

0.74

INC

0.73

VO

0.73

ONY

0.72

iol

0.71

Activations Density 0.010%

terms related to a specific acronym "BB"

references to "BB" in various contexts

No Comments

No Known Activations