INDEX

Explanations

references to public speaking events and interactions with media

oai_token-act-pair · gpt-3.5-turbo

phrases related to legal or formal proceedings

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

hovah

-0.65

 Defin

-0.62

nostic

-0.61

|--

-0.60

 --------------------

-0.59

erenn

-0.58

quartered

-0.57

 revolves

-0.57

 utilizes

-0.54

soDeliveryDate

-0.54

POSITIVE LOGITS

 himself

0.62

 curfew

0.62

 superiors

0.60

sob

0.59

 restraints

0.59

 composure

0.58

 herself

0.58

 deliveries

0.57

 resignation

0.57

 precaution

0.57

Activations Density 1.516%

references to public speaking events and interactions with media

phrases related to legal or formal proceedings

No Comments

No Known Activations