INDEX

Explanations

action-oriented statements related to utilizing features or functionalities

oai_token-act-pair · gpt-3.5-turbo

the modal verb "can" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

pires

-0.72

 Hits

-0.68

breeding

-0.65

bush

-0.64

bats

-0.62

ONSORED

-0.61

ELD

-0.60

 Federation

-0.60

 forthcoming

-0.59

 dwar

-0.59

POSITIVE LOGITS

't

1.47

 optionally

1.30

 choose

1.20

 customize

1.12

 utilize

1.08

 specify

1.08

 omit

1.08

NOT

1.07

use

1.06

 easily

1.05

Activations Density 0.105%

action-oriented statements related to utilizing features or functionalities

the modal verb "can" in various contexts

No Comments

No Known Activations