INDEX

Explanations

phrases indicating specific actions or descriptions typically found in reports or documents

oai_token-act-pair · gpt-3.5-turbo

terms related to formal descriptions and discussions of information or research findings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Scouting

-0.61

cise

-0.60

hunter

-0.59

theless

-0.58

mediated

-0.56

 Decay

-0.56

 Strikes

-0.56

quickShipAvailable

-0.56

USS

-0.55

ipe

-0.55

POSITIVE LOGITS

 therein

1.31

 herein

1.18

by

0.87

 above

0.79

 thereto

0.73

 onstage

0.72

 here

0.72

 below

0.71

 underneath

0.69

 onboard

0.67

Activations Density 0.184%

phrases indicating specific actions or descriptions typically found in reports or documents

terms related to formal descriptions and discussions of information or research findings

No Comments

No Known Activations