INDEX

Explanations

words related to direction or orientation

oai_token-act-pair · gpt-3.5-turbo

phrases indicating direction or movement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

NZ

-0.84

NZ

-0.77

 organisation

-0.74

 organised

-0.73

UK

-0.72

colour

-0.72

 flavour

-0.72

 util

-0.72

 colour

-0.71

 authorised

-0.71

POSITIVE LOGITS

 toward

3.52

 towards

2.68

 Towards

1.88

Tow

1.45

oward

1.37

 upward

1.06

 downward

1.02

 favoring

0.99

 favorably

0.92

 farther

0.89

Activations Density 0.024%

words related to direction or orientation

phrases indicating direction or movement

No Comments

No Known Activations