INDEX

Explanations

names, titles, and locations

oai_token-act-pair · gpt-3.5-turbo

diverse names and identifiers, possibly related to people or characters

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

\<

-0.73

natureconservancy

-0.70

bsite

-0.68

SPONSORED

-0.68

abase

-0.66

 glim

-0.60

etheless

-0.59

lehem

-0.59

userc

-0.58

atin

-0.58

POSITIVE LOGITS

backer

0.90

vous

0.74

issance

0.74

 Janeiro

0.74

 Rouse

0.73

kefeller

0.73

restling

0.73

 Tolkien

0.72

 Grac

0.72

afort

0.71

Activations Density 0.368%

names, titles, and locations

diverse names and identifiers, possibly related to people or characters

No Comments

No Known Activations