INDEX

Explanations

mentions of trees or tree-related activities

oai_token-act-pair · gpt-3.5-turbo

references to trees or forests

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Code

-0.74

Method

-0.73

Witness

-0.72

vol

-0.72

RET

-0.70

Command

-0.70

eu

-0.69

TI

-0.69

ept

-0.68

VI

-0.67

POSITIVE LOGITS

mith

1.14

 trees

1.06

hops

1.05

terday

1.05

ourcing

1.04

 Trees

1.02

ynthesis

0.98

paces

0.98

poons

0.94

hips

0.92

Activations Density 0.049%

mentions of trees or tree-related activities

references to trees or forests

No Comments

No Known Activations