INDEX

Explanations

phrases indicating consequences or outcomes, particularly negative ones

oai_token-act-pair · gpt-3.5-turbo

recurring phrases that indicate caution or warning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Russ

-0.73

estone

-0.61

robe

-0.60

CLOSE

-0.59

onica

-0.57

Fla

-0.56

Enlarge

-0.55

-0.54

roller

-0.54

ione

-0.53

POSITIVE LOGITS

 constitutes

1.21

is

1.09

 seems

1.08

 signifies

1.04

 remains

1.03

 underscores

1.02

 represents

1.00

 proves

1.00

 appears

0.99

 becomes

0.98

Activations Density 0.261%

phrases indicating consequences or outcomes, particularly negative ones

recurring phrases that indicate caution or warning

No Comments

No Known Activations