INDEX

Explanations

phrases related to engaging in particular activities or events

oai_token-act-pair · gpt-3.5-turbo

references to long durations of time or prolonged activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ashtra

-0.72

nil

-0.66

KK

-0.65

fleet

-0.65

illin

-0.64

$.

-0.63

edin

-0.63

Introduced

-0.63

heart

-0.62

mare

-0.61

POSITIVE LOGITS

bidden

1.15

 instance

1.11

 example

1.09

gery

1.06

geries

1.04

gotten

1.03

aging

0.95

cing

0.94

ked

0.93

 starters

0.87

Activations Density 0.480%

phrases related to engaging in particular activities or events

references to long durations of time or prolonged activities

No Comments

No Known Activations