INDEX

Explanations

words or phrases containing "wh"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the sequence "wh"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

PORT

-0.73

phrine

-0.70

 expressing

-0.67

ULAR

-0.66

 mosaic

-0.64

 Comedy

-0.63

ITAL

-0.62

HAEL

-0.61

Reloaded

-0.59

POSITIVE LOGITS

omever

1.32

irling

1.23

ipp

1.16

istle

1.15

olen

1.14

irl

1.14

arf

1.11

ither

1.11

izzard

1.11

ipper

1.10

Activations Density 0.011%

words or phrases containing "wh"

occurrences of the sequence "wh"

No Comments

No Known Activations