INDEX

Explanations

references to places, objects, or concepts related to the word "shade"

oai_token-act-pair · gpt-3.5-turbo

references to varying degrees of shade or color

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

hester

-0.99

raltar

-0.93

inez

-0.85

ITIES

-0.81

sburgh

-0.79

kefeller

-0.77

IJ

-0.75

emort

-0.75

pleted

-0.74

berus

-0.74

POSITIVE LOGITS

 shade

0.97

 shades

0.89

 Shade

0.83

 tint

0.82

="#

0.80

 shading

0.79

 Flavoring

0.77

cliffe

0.76

 curtain

0.74

hue

0.74

Activations Density 0.018%

references to places, objects, or concepts related to the word "shade"

references to varying degrees of shade or color

No Comments

No Known Activations