INDEX

Explanations

texts where the concept of explaining something is emphasized or requested

oai_token-act-pair · gpt-3.5-turbo

instances of the word "explain" and related variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

illet

-0.94

ascus

-0.82

inal

-0.79

estial

-0.77

jet

-0.77

engeance

-0.77

nir

-0.73

ibaba

-0.72

ammy

-0.70

thritis

-0.69

POSITIVE LOGITS

why

1.10

WHY

1.04

 explanations

0.85

why

0.85

ĸļ

0.83

 explan

0.83

 explain

0.82

how

0.82

 explaining

0.81

 explains

0.77

Activations Density 0.034%

texts where the concept of explaining something is emphasized or requested

instances of the word "explain" and related variations

No Comments

No Known Activations