INDEX

Explanations

phrases related to attempted actions or endeavors

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "attempted"

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

minus

-0.75

bin

-0.71

 Origin

-0.71

tal

-0.70

front

-0.68

 Foot

-0.68

Experience

-0.68

quality

-0.68

 Wrap

-0.67

eyes

-0.67

POSITIVE LOGITS

 attempted

1.06

 attempts

1.05

 unsuccessfully

1.03

Attempts

1.01

 attempt

1.00

 carbohyd

0.98

Attempt

0.90

urdue

0.90

gomery

0.87

 Attempt

0.85

Activations Density 0.008%

phrases related to attempted actions or endeavors

occurrences of the word "attempted"

No Comments

No Known Activations