INDEX

Explanations

numerical ranges or categories within text

oai_token-act-pair · gpt-3.5-turbo

numerical data or statistical information

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 lull

-0.69

 Federation

-0.67

 Leilan

-0.66

 hostage

-0.64

 blot

-0.64

 coales

-0.63

irds

-0.63

 kidn

-0.60

Bul

-0.59

 negotiations

-0.58

POSITIVE LOGITS

external

0.86

requires

0.84

âĢ¦)

0.79

(/

0.79

?)

0.78

!)

0.78

NOT

0.75

pdf

0.74

olitan

0.74

(*

0.73

Activations Density 0.008%

numerical ranges or categories within text

numerical data or statistical information

No Comments

No Known Activations