INDEX

Explanations

possessive pronouns followed by a word describing a state or condition

oai_token-act-pair · gpt-3.5-turbo

possessive pronouns and references to ownership

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

bender

-0.65

FTWARE

-0.61

umeric

-0.61

apter

-0.61

ocument

-0.61

bill

-0.61

hemy

-0.60

gans

-0.59

ption

-0.59

vation

-0.58

POSITIVE LOGITS

 expense

1.32

 behest

1.29

 fingertips

1.20

 doorstep

1.19

 disposal

1.13

 discretion

1.09

own

1.00

 insistence

0.98

 leisure

0.95

 helm

0.94

Activations Density 0.043%

possessive pronouns followed by a word describing a state or condition

possessive pronouns and references to ownership

No Comments

No Known Activations