INDEX

Explanations

phrases related to negative emotional states

oai_token-act-pair · gpt-3.5-turbo

statements about occurrences or conditions

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

poke

-0.72

 races

-0.66

resp

-0.65

 Write

-0.65

wrote

-0.62

 WRITE

-0.62

 learns

-0.62

 Supports

-0.62

itely

-0.61

 Classes

-0.60

POSITIVE LOGITS

 exacerbated

1.20

 contrasted

1.08

 attributable

1.04

 outwe

1.04

 evident

1.04

 sympt

1.02

 contagious

1.02

 overshadow

1.01

 reflected

1.00

 compounded

1.00

Activations Density 0.377%

phrases related to negative emotional states

statements about occurrences or conditions

No Comments

No Known Activations