INDEX

Explanations

pronouns for people and actions related to them

oai_token-act-pair · gpt-3.5-turbo

pronouns indicating subjects engaged in actions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Effective

-0.70

Reviewer

-0.68

acha

-0.67

ãĥĳ

-0.65

 exception

-0.61

 sooner

-0.61

 banning

-0.61

Asset

-0.61

 overriding

-0.61

quickShipAvailable

-0.61

POSITIVE LOGITS

 awaited

1.15

 approached

1.12

 prepares

1.09

 walked

1.09

 waited

1.09

 slept

1.07

 watched

1.05

 exited

1.04

 struggled

1.04

 marched

1.03

Activations Density 0.083%

pronouns for people and actions related to them

pronouns indicating subjects engaged in actions

No Comments

No Known Activations