INDEX

Explanations

names or titles within sentences

oai_token-act-pair · gpt-3.5-turbo

phrases that reference names or possessive constructs

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

iasm

-0.79

venants

-0.75

issions

-0.72

phasis

-0.71

 grasp

-0.70

istas

-0.67

 crumble

-0.66

ienne

-0.64

agy

-0.63

ista

-0.62

POSITIVE LOGITS

 Uzbek

0.84

 culprit

0.72

ãĤ·ãĥ£

0.71

 Killer

0.66

BOOK

0.65

 supplier

0.64

 suspects

0.63

ãĤº

0.63

Tup

0.62

pages

0.61

Activations Density 0.116%

names or titles within sentences

phrases that reference names or possessive constructs

No Comments

No Known Activations