INDEX

Explanations

numerical and alphabetic patterns within specific contexts

oai_token-act-pair · gpt-3.5-turbo

references to announcements and notifications related to various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

jet

-0.81

fair

-0.78

tons

-0.69

aug

-0.67

plain

-0.66

jun

-0.66

kell

-0.66

fal

-0.65

comb

-0.65

boy

-0.65

POSITIVE LOGITS

ING

1.77

IES

1.77

ITY

1.77

ABLE

1.70

ED

1.69

ATIONS

1.69

IZ

1.69

ION

1.68

ATION

1.66

ORS

1.66

Activations Density 0.146%

numerical and alphabetic patterns within specific contexts

references to announcements and notifications related to various topics

No Comments

No Known Activations