INDEX

Explanations

phrases related to individuals or locations with "iw" in them

oai_token-act-pair · gpt-3.5-turbo

occurrences of a specific term related to technological or collaborative contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dent

-0.69

Express

-0.65

 Lamar

-0.64

 Spears

-0.62

 Baal

-0.62

 Arena

-0.61

 Abdel

-0.61

 trauma

-0.60

 elimination

-0.60

 stops

-0.59

POSITIVE LOGITS

iw

4.67

iband

1.26

iam

1.22

iy

1.19

isu

1.17

iu

1.14

ew

1.11

kj

1.06

ih

1.05

wcsstore

1.05

Activations Density 0.012%

phrases related to individuals or locations with "iw" in them

occurrences of a specific term related to technological or collaborative contexts

No Comments

No Known Activations