INDEX

Explanations

mentions and references to the television network "FOX"

oai_token-act-pair · gpt-3.5-turbo

references to the FOX network

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

agn

-0.79

ucket

-0.77

achy

-0.74

assy

-0.73

idences

-0.71

anks

-0.70

haps

-0.70

anked

-0.68

ivari

-0.68

ampunk

-0.66

POSITIVE LOGITS

FOX

1.20

conn

1.03

FOX

0.92

çīĪ

0.80

fox

0.79

 Fest

0.76

DEV

0.75

Fox

0.74

bane

0.71

CAST

0.70

Activations Density 0.007%

mentions and references to the television network "FOX"

references to the FOX network

No Comments

No Known Activations