INDEX

Explanations

names of a specific individual, "Jake"

oai_token-act-pair · gpt-3.5-turbo

the name "Jake" across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 conduc

-0.78

iary

-0.75

iated

-0.73

amera

-0.65

acent

-0.64

 subst

-0.63

arily

-0.62

oppable

-0.62

Ħ¢

-0.62

Ü

-0.62

POSITIVE LOGITS

glers

1.00

Gy

0.89

ansas

0.88

 Jake

0.83

unin

0.83

 Skywalker

0.81

 McGee

0.77

EStream

0.76

cki

0.76

caster

0.75

Activations Density 0.021%

names of a specific individual, "Jake"

the name "Jake" across various contexts

No Comments

No Known Activations