INDEX

Explanations

and recognize mentions of physical or metaphorical "ticks" or "ticking"

oai_token-act-pair · gpt-3.5-turbo

terminology related to checking off or marking items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ocally

-0.72

pite

-0.71

 ILCS

-0.66

english

-0.66

ufact

-0.64

uthor

-0.62

 Galile

-0.62

perty

-0.61

ocal

-0.61

ãĥ´ãĤ¡

-0.61

POSITIVE LOGITS

lish

1.04

ety

0.99

ched

0.94

ering

0.94

lers

0.94

leness

0.90

 tick

0.85

ling

0.84

tick

0.82

boxes

0.81

Activations Density 0.017%

and recognize mentions of physical or metaphorical "ticks" or "ticking"

terminology related to checking off or marking items

No Comments

No Known Activations