INDEX

Explanations

references to glass or glass-related objects

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "glass."

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

merce

-0.73

IRED

-0.73

ilitary

-0.73

unct

-0.72

CVE

-0.68

ften

-0.65

Express

-0.65

DAY

-0.64

Filename

-0.64

 Cosponsors

-0.64

POSITIVE LOGITS

glass

1.09

 shards

1.01

 bottles

0.97

ware

0.96

 beads

0.96

 glass

0.94

works

0.93

caps

0.91

 bead

0.90

 pane

0.88

Activations Density 0.011%

references to glass or glass-related objects

mentions of the word "glass."

No Comments

No Known Activations