INDEX

Explanations

terms related to transparency

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

examples of the word "transparent"

oai_token-act-pair · gpt-3.5-turbo

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

ander

-0.72

skill

-0.72

ARK

-0.69

 Upton

-0.68

Ir

-0.67

 Braun

-0.67

 Extrem

-0.65

akings

-0.64

 Lower

-0.64

req

-0.64

POSITIVE LOGITS

 transparent

1.35

 transparency

0.96

parency

0.96

 opaque

0.96

 translucent

0.93

 opacity

0.83

iary

0.81

ewater

0.81

ible

0.80

recy

0.79

Activations Density 0.005%

terms related to transparency

examples of the word "transparent"

No Comments

No Known Activations