INDEX

Explanations

specific nouns related to physical objects or entities, such as "disk," "bike," "gun," and "jar."

oai_token-act-pair · gpt-3.5-turbo

notable entities or significant items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 policies

-0.72

afi

-0.65

 absentee

-0.64

 salaries

-0.63

 provinces

-0.63

essions

-0.62

ional

-0.62

 Policies

-0.61

 principals

-0.60

respons

-0.60

POSITIVE LOGITS

washer

0.77

Rot

0.69

 Doodle

0.64

Redditor

0.63

 crochet

0.63

eus

0.62

 Pluto

0.61

 flown

0.60

 slic

0.60

Recipe

0.60

Activations Density 0.874%

specific nouns related to physical objects or entities, such as "disk," "bike," "gun," and "jar."

notable entities or significant items

No Comments

No Known Activations