INDEX

Explanations

words related to investigating, reviewing, evaluating, assessing, and responding to various types of issues or situations

oai_token-act-pair · gpt-3.5-turbo

phrases that reflect ongoing actions or processes

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

hov

-0.73

 Powered

-0.73

lav

-0.72

owered

-0.70

uel

-0.68

unts

-0.67

©¶æ¥µ

-0.67

 Slug

-0.67

 Golf

-0.64

 Pinball

-0.64

POSITIVE LOGITS

 evaluate

1.17

 respond

1.13

 quantify

1.06

 remed

1.05

 critique

1.05

 analyze

1.04

 corrections

1.02

 mitigation

1.02

 formulate

1.01

 analyse

1.01

Activations Density 0.198%

words related to investigating, reviewing, evaluating, assessing, and responding to various types of issues or situations

phrases that reflect ongoing actions or processes

No Comments

No Known Activations