INDEX

Explanations

words related to awakening or becoming alert

oai_token-act-pair · gpt-3.5-turbo

gerunds and present participles associated with various actions and conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 torn

-0.72

 atro

-0.65

 thor

-0.63

 tears

-0.61

bul

-0.60

 lucky

-0.59

 Weasley

-0.58

 riot

-0.58

 hungry

-0.58

 nozzle

-0.58

POSITIVE LOGITS

olate

0.93

akes

0.92

ESE

0.85

ernel

0.79

umar

0.78

DOM

0.77

VID

0.77

arta

0.76

hran

0.76

velop

0.75

Activations Density 0.012%

words related to awakening or becoming alert

gerunds and present participles associated with various actions and conditions

No Comments

No Known Activations